ROCm/composable_kernel
1
0
Fork 0
mirror of https://github.com/ROCm/composable_kernel.git synced 2026-05-03 05:01:25 +00:00
Code Issues Packages Projects Releases Wiki Activity

Add int4 example for convnd_fwd_bias_relu_add (#375)

Browse Source 
* Add int4 example for convnd_fwd_bias_relu_add

* Fix AddReluAdd for building without int4 support

* Update CMakeLists.txt

* Format

* Convert int4 tensors for int8 kernel

* Fix device memory allocation

* Format

* Format
This commit is contained in:
Rostyslav Geyyer
2022-08-25 17:08:43 -05:00
committed by GitHub
parent d520d0cfc1
commit b73ae24234
8 changed files with 665 additions and 107 deletions
Download Patch File Download Diff File Expand all files Collapse all files

12
include/ck/tensor_operation/gpu/element/element_wise_operation.hpp
View File

@@ -98,6 +98,18 @@ struct AddReluAdd
int32_t c = b + x2;
y = c;
}
#ifdef CK_EXPERIMENTAL_BIT_INT_EXTENSION_INT4
template <>
__host__ __device__ constexpr void operator()<int4_t, int8_t, int4_t, int4_t>(
int4_t& y, const int8_t& x0, const int4_t& x1, const int4_t& x2) const
{
int32_t a = x0 + x1;
int32_t b = a > 0 ? a : 0;
int32_t c = b + x2;
y = c;
}
#endif // CK_EXPERIMENTAL_BIT_INT_EXTENSION_INT4
};
struct AddHardswishAdd